Quantitative analysis of smooth pursuit eye movement

Abstract

Abnormalities of smooth pursuit eye movement (SPEM) have been estimated, mainly using the wave form on an electro-oculogram, in a qualitative way. Many methods for quantitative analysis of SPEM have been designed, though most are still uncommon in present clinical use. Using a personal computer, we developed a method of automatic quantitative analysis of ocular tracking eye movement recorded by electro-oculography (EOG). The design concept of this method is based on the observation that eye movement during ocular tracking consists of two different kinds of eye movements, one is SPEM and the other is saccade. The combination of SPEM and saccade (composite eye movement: CEM) commonly appears during ocular tracking. These two kinds of eye movement are essentially different not only in behavior but also about involved neural pathway in the central nervous system. From this point of view, we believe that the two kinds of eye movements involved in ocular tracking should be evaluated separately. The analysis method is outlined as follows. A horizontal sinusoidally moving visual target was employed to elicit ocular tracking eye movements. The test frequencies were set at 0.1, 0.2, 0.4 and 0.8Hz, and the amplitude of target motion was 15 deg at each frequency. The 20 seconds of eye movement data measured by EOG were fed into the computer through a digital-analog converter for further analysis. Using our original saccade detection algorithm, based on the physiological behavior of saccades, the saccadic components were detected and removed from the eye movement wave. The remaining parts, fragments of SPEM, were connected by means of interpolating defective parts. The reconstructed wave was a slow cumulative eye position curve (SCEP). Sinusoidal target motion, CEM and SCEP were processed by the FFT (Fast Fourier Transformation) method. Bode plots were applied to summarize the gain and phase of responses to SCEP and the target motion wave. These processes enable us to estimate abnormalities of SPEM such as low gain, abnormal phase shift and large trends in tested duration. We conclude that the method described here is useful for quantitative estimation of SPEM in clinical neuro-otological examinations.